SarathChandra Mandadi

Technical Lead, AI/ML Development — Bosch Global Software Technologies

Hyderabad · October 2017 – February 2026

• Led end-to-end implementation and solution delivery of AI/ML, NLP, Deep Learning, Knowledge Graph, and Generative AI applications for Bosch engineering teams — owning technical architecture, cross-functional execution, mentoring, roadmap planning, and stakeholder communication while delivering 20–60% efficiency, effort, and cost savings through intelligent automation
• Designed and productionized AI applications and microservices using Python, FastAPI, Docker, Azure services, Streamlit, and React, translating research-led models into scalable internal products with stronger maintainability, reliability, and adoption
• Partnered with business stakeholders, product teams, data teams, and academic collaborators (IISc Bangalore, IIT Madras, IIT Hyderabad, IIT Tirupati) to define AI roadmaps, benchmark model performance, and deliver enterprise-grade AI solutions across the full lifecycle

Key AI/ML Projects Delivered at Bosch

• Requirement Similarity Assist (2022–2026): Architected a Generative AI and NLP platform for comparing current and historical stakeholder requirements using BERT-based semantic similarity, LLM reasoning, HyDE retrieval, agentic workflows, LangChain/AutoGen-style orchestration, FastAPI microservices, React, Azure, and vector databases (FAISS, Pinecone, LanceDB) — driving 20–40% savings in analysis effort and turnaround time
• Req2Test / Testcase Generation Assist (2018–2026): Built an AI-driven test case generation solution using Neo4j knowledge graphs, requirement relation extraction, RAG pipelines, LLM-assisted generation, and cloud-hosted Python APIs to derive structured test assets (pre-conditions, actions, post-conditions, positive/negative scenarios), reducing manual authoring effort by ~30%
• Requirement Intelligence Suite (2017–2021): Designed NLP and Deep Learning applications for requirement ambiguity detection, coreference resolution, relation extraction, contradiction-entailment detection, summarization, and formalization using CRF, HMM, LSTM, Bi-LSTM, Bi-LSTM-CRF, BERT, PyTorch, spaCy, Stanford NLP, AllenNLP, T5, FLAN, and UL2 — delivering 20–35% efficiency gains in requirement review workflows
• Knowledge Graph for Requirement Mining and Traceability (2019–2021): Implemented a reference architecture for graph-based requirement intelligence by extracting entities, relations, and triplets from Bosch proprietary engineering documents and IBM DOORS data into Neo4j, enabling graph querying, dependency tracing, and downstream automation
• Automotive Manual RAG Assistant and Context Chatbots (2021–2023): Built GenAI-powered chatbot and contextual Q&A systems over Bosch automotive manuals using BERT, BiDAF, GPT-family models, LLaMA, Mistral, Mixtral, FastAPI, Streamlit, vector retrieval, and Azure application services — reducing manual document lookup and support effort by 20–40%
• Multi-Agent Document Intelligence Platform (2022–2026): Designed multi-agent AI applications for document querying, clause retrieval, document similarity, and multimodal extraction of text, tables, and images from PDFs using Python, FastAPI, React, Azure, vector databases, and retrieval pipelines — contributing to 20–50% savings in manual document analysis effort
• Vehicle Trace Catalogue (2019–2021): Developed an ML/NLP-based analytics solution to process Sweden vehicle drive-test trace logs and classify pass/fail outcomes against expected driver instructions using Python, Hugging Face, and BERT-based models
• VivaRay and VivaScope (2019): Delivered applied ML and Deep Learning solutions in healthcare — regression-based anemia classification from optical sensor data and CNN-based microscopic image analysis for cell abnormality detection
• Defect Prediction, Bug Localization, and Topic Modeling (2020–2022): Built ML pipelines for software defect prediction, bug classification, clustering, and trace-path localization, improving engineering investigation productivity by up to 40% in selected workflows
• Model Evaluation, Benchmarking, and Fine-Tuning Framework (2018–2026): Evaluated and benchmarked BERT models, transformer models, and LLMs, and fine-tuned selected models using PEFT, LoRA, and QLoRA for Bosch AI applications; established ground-truth-based evaluation workflows for fit-for-purpose deployment
• AI Platform Delivery and Team Leadership (2017–2026): Led distributed teams across AI streams, defined epics, user stories, and solution roadmaps, mentored engineers and researchers, and drove adoption of application architecture, API-first design, CI/CD-aligned delivery, and containerized deployment

Natural Language Modelling Engineer — Unisys Corp

Bengaluru · April 2017 – August 2017

• Designed, developed, and fine-tuned NLP models for text classification, sentiment analysis, and language understanding use cases
• Trained models on large, enterprise-scale text datasets using ML and deep learning techniques, optimizing architectures and hyperparameters for prediction quality and runtime efficiency
• Implemented end-to-end NLP pipelines covering data preprocessing, feature extraction, model training, evaluation, and inference
• Collaborated closely with software engineering teams to integrate NLP models into production systems, ensuring scalability, robustness, and reliability in enterprise environments

Senior Data Research Analyst — Avaamo Technologies

Bengaluru · December 2015 – October 2016

• Designed and developed enterprise conversational AI chatbots using Google Dialogflow (API.ai), Facebook Wit.ai, Amazon Lex, and Microsoft LUIS
• Built and optimized NLU pipelines including intent classification, entity extraction (NER), and slot filling; refined decision trees, dialog flows, and fallback strategies to improve chatbot accuracy
• Applied NLP techniques to analyze unstructured data (user conversations, application reviews), performing expectation mining and sentiment analysis to surface user needs and feature gaps
• Supported bot onboarding and training on domain-specific enterprise data and contributed to scalable chatbot deployments for customer support and enterprise automation

Senior Market Research Analyst — Alten-Calsoft Labs

Bengaluru · April 2015 – December 2015

• Performed customer segmentation and market analysis using clustering techniques to identify behavioral patterns and target segments
• Applied text mining and sentiment analysis on large-scale Twitter data using R to capture customer perception and market trends
• Conducted competitor analysis and market intelligence research focused on digital transformation initiatives
• Translated analytical findings into actionable insights for senior stakeholders including VPs of Marketing and Directors

Research Analyst, Data & Global Operations — FactSet Research Systems Inc

Hyderabad · July 2011 – March 2015

• Analyzed financial documents (Annual Reports, SEC filings — 10K, 4K, win–loss data) to extract business and competitive insights
• Extracted CCPS entities (Customers, Competitors, Partners, Suppliers, Peers) from unstructured financial text for downstream analytics
• Implemented early NLP techniques for Named Entity Recognition using Stanford CoreNLP with CRF models, combined with rule-based extraction tools (AutoIt, ImportIO, Hxcel) to improve accuracy
• Performed text mining, sentiment analysis, and clustering on lead/prospect comments to support investment and sales intelligence
• Built AI-driven dashboards and visualizations, including world map visualizations of global entities and linked relationship graphs

M.Sc. (Chemistry) — Osmania University

Hyderabad · August 2008 – June 2010

B.Sc. (Biotechnology) — Kakatiya University

Telangana · August 2006 – June 2008

Professional Certifications

• DELL EMC Data Science Associate - Proven Professional certification
• Azure Data Factory For Data Engineers - Udemy
• Deep Learning (Andrew Ng) - Coursera
• Python Machine Learning & Data Visualization - Udemy
• Statistics with Excel - Lynda.com